While waiting for Peter Bol’s talk to begin I can’t help but think that this massive project will be decades to complete. The amount of data – China has been keeping censuses for over 2000 years – is staggering, and GIS is hardly the easiest software to manipulate. Professor Will Thomas began the morning session introducing Bol and his work. Behind him, on the screen was the CHGIS website: www.fas.harvard.edu/~chgis. Thomas explained that UC-Berkeley was sponsoring CHGIS, adding that Bol’s work was “crucially important,” and Bol professionally is a “great practitioner of digital history.”
Bol says that he is here to explain the basics of and hint at the high-end possibilities of CHGIS, but noted that. “I’m not going to try and match Ed Ayers from last night.” Who would try? Ayers’s talk last night was captivating, inspiring, provocative, but considering the ambitions of CHGIS, perhaps Bol is being a little too modest.
Bol displays a late sixteenth century map of China. He explains that in paper maps multiple kinds of data are displayed, but the data exists in the aggregate. GIS dis-aggregates map, meaning that the user can strip or add information and more closely examine the source data. CHGIS says, “let’s treat it [data] as layers.” Bol adds layers to the sixteenth century map and rivers appear magically. The map displays how political history in China worked. County seats invaribly emerged on rivers, while higher-end administrative centers were placed on junctions of the same rivers. Graphically, CHGIS reveals the geographic structure of political system in sixteenth century China in ways a manuscript cannot. The information came, unlike in a book or paper, with a simultaneity that was impressive.
Next Bol brings up a map of western China generated by GIS. The screen fills with numbers. Numbers everywhere, and it strikes one as a bit overwhelming. Bol is quick to remind us that a historical GIS map is not a map, per se, but rather a database. This must be kept in mind, he cautions, otherwise we will find that the best features of GIS will be obscured.
Bol shows us what he means and displays a giant database of Chinese counties and then the map the raw data subsequently generates. Again, the map is extremely busy, but I suppose there is a zoom function that would allow the interested scholar to focus in on an area of interest. The database, Bol explains, allows one to take different moments in time and compare them in ways. The technology allows the scholar to make connections that would otherwise be obscure. For example, Bols shows us prefectural boundaries in 1102 compared to prefectural boundary in 1550s. He then shows the same map, but this time with county seats, ranging from 1086 to 1550. County seats, however, do not disappear, but they get added, map shows. Why? Bol cannot say, but notes that it is an interesting question, one that would not likely have been raised without CHGIS.
Bol’s shows what he claims to be one of the most important maps in human history – the earliest grid, carved into twelfth century stone. We don’t see the real thing, but like everything else, see its image (or rather data that represents its image) on the screen. The great thing about this map or any map, according to Bol, is that it represents something that no one else can see. Maps represent that landscape beyond what anyone can experience. They are not reality, but rather arguments; attempts to convince others, the emperor for example, that this was what China really is.
Bol riffs: “Oh by the way emperor, this is what the empire you rule over looks like.” It’s a provocative point, one that Bol will return to later. Maps, as I concluded an enterprise of human imagination. We imagine what China, or anywhere else ought to look like. To me, this does not render illegitimate maps or the CHGIS project, but is rather another statement of how humanists get at truth. We are not scientists who believe that we are just trying to describe the reality that’s “out there” in the objective world, but we instead use our creative and intellectual powers to pierce through a potentially deceptive reality to find a truth.
Bol turns next to tracing the changing paths of rivers through time. This endeavor is essentially a history of Chinese maps that also makes use of archeological data, and historical texts. The example shows how interdisciplinary and intellectually challenging digital map making can be.
Bol makes a bold claim, stating that what he and his colleagues are creating not only will be but already is more authoritative than any print atlas. If you want authority, you have to turn to the GIS. Now, it is generally understood that CHGIS has replaced any and all print atlases. demonstrating the power, potentially of digital history and geography.
Now the discussion turn to a GIS question that I’m more familiar with: does one simply create a database or try for something “more.” How, in other words, does a digital scholar take advantage of the web to exploit the peculiar features of this medium. Bol comes down on the the second option. He says that to fully utilize the collaborative nature of the web, that he sees the CHGIS as an infrastructural GIS project that provides a base to allow other users to build on work you’ve done. Bol namechecks Will Thomas and me in this regard, and I enjoy being pointed out. We, along with a whole group of others, are trying to do “something more” with the Railroads and the Making of Modern America. Bol offers all the CHGIS databases for free so that anyone can join their data to his stuff. This sounds like a true community of scholars via digital work.
Another feature that makes CHGIS and GIS projects different from their print counterparts is that the databases allow one to exhibit all of the sources used to create the map. All primary sources used are documented and available for inspection. Geographers’s big problem, says Bol, is that they give you the map and say “believe me!” Bol gives you he sources. You can check him out and verify for yourself. Perhaps the humanities are becoming more like the sciences, allowing users to reconstruct our work and replicate it, like a scientific experiment. Still, this does not in any way refute Bol’s earlier assertion about maps being a product of human imagination, informed, of course, by data.
Bol, like everyone else, is impressed by Google maps.
China not only possesses about 20 percent of the world’s population, but has also had a bureaucratic organization that has complied data for thousands of years on just about everything that population did, which means the amount of data available is staggering. China is a great HGIS candidate. But what can one use CHGIS to learn from these ancient bureaucrats’ scrupulously-kept records? Bol picks South China an example, pointing to Guangdong’s experience. For mysterious reasons, the official record shows that over time, there were fewer county seats in the province, contradicting other data that shows China’s population doubling from 750-1050. This is a historical problem. Bol initially thinks the problem to be a local rebellion. Guangdong was a great commercial center and rebellion targets foreign merchants. The rebelling kills 100,000 merchants thus wiping out the provinces main source of wealth. The local government, however, was withering before all this violence and commercial suicide. The answer is perhaps obscure but the CHGIS gives Bol and the rest of us an intriguing question. I guess the questions are more important than the answers; others can attack those.
Now Bol displays a map that shows the divergence of the economic and the government hierarchy in the 11th century. It is, the most important event in Chinese history and because of CHGIS it is now graphically represented.
The best part of the CHGIS, according to me, is the biography database. On the screen appears a timeline, showing the number biographies completed in specific historical eras. This raises a whole bunch of historical questions, such as: when were biographies were written and where; what historical events lead to an increase in biography? I would like to say that those officials living in a time of declension would be more apt to document their times, the fall. But I’d have to test that against CHGIS data. Biographies form backbone of narrative history, but certain times have a glut and others a scarcity. This is significant because it distorts history. The biography database allows us to recognize this and work to correct it.
Bol has only given two more minutes and spends it discussing the relative merits of polygons. Much of Bol’s explanation is beyond me, but I think I get the idea that the boundaries are artificial, imaginary, and potentially misleading.
Questions from the audience (only one): Can you show significances not immediately visible? Bol, being facetious, notes about digital mapmaking: “Isn’t it great that we can lie, but now we can lie in diverse ways.” Maps allow us to lie because we create things outside our experience, that are essentially products of our imagination, not replicas of reality, so one should be careful making and using maps. I agree, but would add the caveat that lying and subjective truth are not that far apart. For artists, lying is a great thing. The most creative among us are often the best liars. Akira Kurosawa’s Rashomon explores the issues of lying, storytelling, creativity, perspective, and truth much better than I can here. Then Bol turns to American history and argues that race and class are less variable than spatial narratives in U. S. history, meaning that where you live in a city is more important than issues of race or class in public health issues. Now that’s a provocative point that’s just left hanging there. I guess it’s sort of out of the range of this discussion, but boy. How about we combine race and space? I would think that race (and class too) structures spatial organization in U. S. cities.
Patrick Manning appears behind the lecturn and its on to the next presentation.